Open Teaching
A Case Study on Publishing Lecture Videos Publicly
Richard Buckland
School of Computer Science and Engineering
The University of New South Wales
Sydney 2052 Australia
Email: richardb@unsw.edu.au
Abstract
This paper reports on a project to video record lectures for undergraduate computing courses and publish them openly online. It outlines the objectives
of the project, the techniques experimented with to
record the lectures, the major challenges encountered,
and how they were addressed. The paper then highlights a number of the interesting and sometimes unexpected benefits which have followed from publishing
the lecture recordings, as well as some disadvantages
and warnings. It is my hope that others may be able
to benefit from our experiences and feel empowered
to record and openly publish their own teaching practice.
Keywords: OpenLearning open data online lecture
video recording eLearning YouTube
1
Introduction
“The internet is fantastic. I am receiving a
university-level education whilst I sit here in
my underwear in my mom’s basement, eating nachos and drinking coke. Life is good.”
2
YouTube user caesiume commenting on video ‘Lecture 1: Introduction to Data Structures and Algorithms’ (caesiume 2010)
This paper reports on the experiences of a project
to video record undergraduate lectures and publish
them openly online. In 2008, with the encouragement
and enthusiastic support of former students, I started
capturing lectures of our first and second year computing courses on video. I delivered the lectures and
the former students operated the cameras and live
mixing equipment. My original intention was merely
to make the recordings available to a handful of offcampus students and the undergraduate teaching academics within our school (CSE, the School of Computer Science and Engineering). However in the spirit
of openness in education and after much agonising I
also made the recordings publicly and freely available
to all via YouTube and iTunes. In the subsequent two
years these recorded lectures have been viewed over
Links and supplementary material for this
https://wiki.cse.unsw.edu.au/richardb/ACE2011
paper
a million times by hundreds of thousands of learners of widely varying age, nationality, socio economic
background, and prior knowledge and experience.
Since that time there has been a constant demand
via email, YouTube comments, and physical mail for
further such recordings and for supplementary material to support the current recordings. There has
also been a similar volume of messages of thanks and
gratitude from learners and fellow teachers worldwide
for sharing the lectures. I have been delighted to find
that what was originally envisioned as a small project
has blossomed into one of the most significant educational activities I have undertaken.
This paper outlines on the techniques we experimented with to record the lectures, the major problems we encountered, and how we have addressed
them. The paper then highlights a number of the
interesting and sometimes unexpected benefits which
have followed from publishing the lecture recordings,
as well as some disadvantages and warnings. It is my
hope that others may be able to benefit from our experiences and feel empowered to record and openly
publish their own teaching practice.
at
c
Copyright 2011,
Australian Computer Society, Inc. This paper appeared at the Thirteenth Australasian Computing Education Conference (ACE2011), Perth, Australia, January 2011.
Conferences in Research and Practice in Information Technology (CRPIT), Vol. 114, John Hamer and Michael de Raadt, Ed.
Reproduction for academic, not-for profit purposes permitted
provided this text is included.
Background
The initial impetus to video record first year CS1
computing lectures in 2008 arose from two problems
CSE was facing at that time. Firstly we wished to
offer selected secondary school students the chance to
take a first year university computing course whilst
still at secondary school but recognised they would
find it hard to miss school and travel to campus several times each week to attend lectures. Secondly
we wished to share teaching practice amongst the
pool of teachers for our core first and second year
courses to both demonstrate how a new and ambitious syllabus could be implemented, and to ensure
that course coverage remained relatively stable from
semester to semester.
2.1
For viewing by students
Video recording lectures was a key part of our secondary school outreach program. Secondary school
students in the programme were to watch the lecture
videos in their own time in advance and then attend
university once each week for a face-to-face tutorial
and lab. In 2008 we ran a trial with three secondary
school students, we expanded this to sixteen in 2009,
and forty in 2010.
Critical requirements: To be able to have the
secondary school students participate in the course in
synchrony with our undergraduate students the lecture recordings needed to be available to the participants within a few days of the actual lecture - ideally
the next day for the Thursday lectures — since the
exercises for week n + 1 week depended on the lecture material from week n, and many students liked
to prepare on the weekend. Furthermore it was imperative that the end-to-end process be reliable and
robust - if the lecture recording process ever failed for
any reason there was no easy way for the students to
recover, that lecture would be missed.
Desirable requirements: We wanted the remote secondary school students to feel that they were
part of the course community, and that the undergraduate students taking the course were their peers.
Our objective for the lecture recordings was that
the viewer would feel that they were sitting in the
class. To this end we used a Cinéma Vérité approach
to the filming. This precluded using a simple screencapture plus static camera arrangement such as the
proprietary Lectopia system which was then being
used to audio record some lectures at UNSW.
Instead we used a human camera operator
recorded each lecture from the back row of seats. The
operator was instructed to point the camera wherever
they felt that they would be looking if they were a
student in the class and to zoom on things that they
would want to see more closely. We didn’t use experienced camera operators, instead we used enthusiastic
ex-students of the course. We hoped that this would
mean they would have some insight into the mind of
the student viewers.
We did not feel that elaborate post production was
of key importance, or even that the video quality was
critical, so long as it was sufficient that the students
were able to see and understand what was going on.
In summary we strove to give the recordings a sense
of authenticity rather than a sense of professional production.
2.2
For viewing by teachers
The second factor motivating our venture into lecture
recording was to enable the teachers of our core first
and second year courses to share their teaching practice, and to stabilise the content of these courses.
In 2006 CSE’s core syllabus was reviewed and substantially restructured. Before the review and restructure first and second year courses had learning
objectives predominantly in the lower half of Bloom’s
taxonomy (Bloom 1956) (remembering, understanding, applying in the terminology of Anderson (2001)).
The content objectives of each course were almost exclusively low level discipline knowledge and skills. In
contrast the new syllabus added a sequence of higher
level learning objectives and explicitly included the
development of graduate attributes in each course.
For example, after the restructure developing effective group-work, leadership, innovation, skepticism,
rigour, time management, life long learning skills,
and a framework of ethics and professionalism are
now objectives of each course rather than simply being addressed at the end of the degree program in a
standalone fashion — for example by a single final
year course dedicated to group-work and time management.
However in the period immediately following the
restructure a number of the pool of teachers who
taught these core courses expressed uncertainty as to
how this integrated approach could be achieved in
their course, or even skepticism that it were possible
to do so at all in more than a token way. I felt that in
addition to talking about ways the new syllabus could
be implemented it could be useful and compelling to
directly show concrete examples of ways of achieving
these objectives — and to that end over the past two
years I have run, video recorded, and published a reference version of each of these core first and second
year courses.
A related challenge which the school wanted to
address was the issue of syllabus drift. Each of our
large core courses is taught by at least two teachers —
one in first semester and one in second semester. Furthermore over time teachers move off courses and new
teachers take their place. We have observed that individual teachers have their own understanding and
interpretation of the syllabus and that the material
covered in a course, and the depth to which it is
covered, can change substantially from year to year,
even from semester to semester. This can cause problems in subsequent courses which rely on the material
which was supposed to be covered. For a personal example: in my Computer Security and Cryptography
course in 2008 I had students who had never heard
of the RSA cryptosystem, students who had already
been taught it once, students who had been taught it
twice, and even some outraged students who, because
of the particular sequence of semesters in which they
had taken the prerequisite courses, had already seen
it three times!
The school has instituted a series of measures to
limit the problem of syllabus drift. Having a public
and published reference version of each course is an
important part of our approach.
3
Publishing Online: Lofty Goals
The previous section described my initial motivation
for video recording lectures. In this section I discuss the decision to then publish those recordings
online, free for all to access and view. In the Outcomes section to follow I will posit that this decision
to publish openly, mirroring the emerging movement
for open data generally (see for example (Uhlir &
Schroder 2008), (OECD 2004), and the data.gov and
data.gov.uk sites) was the most important decision
of the project. As MIT President Charles M. Vest,
whose inspirational OpenCourseWare (OCW) project
pioneered open teaching, proclaimed “[openly sharing
lectures] expresses our belief in the way education can
be advanced - by constantly widening access to information and by inspiring others to participate.” (Vest
2001)
The initial goals for undertaking the recording
were modest and addressed pragmatic issues at my
own institution. In contrast my hopes for publishing
the material online were more lofty — to disseminate discipline content and practices to categories of
learners traditionally unable to access universities for
learning, and to encourage improvements in the professional practice of teaching.
3.1
Openness and Autonomy
Openly publishing lectures has the potential for
widespread improvement in the practice of teaching
by introducing a novel form of openness into the
teaching process.
What happens in the classroom has traditionally
been a private matter between teacher and students.
Referring to Britzman (1986)’s observations on the
culture of teachers Rogers & Babinski (2002) posit:
‘It is this “culture of teachers” which promotes privacy and autonomy, which establishes barriers to genuine dialog among
teachers.’ and ‘there are invisible walls constructed by the “culture of teachers” that
perpetuate a lack of community’
Afterwards, in some cases, some aspects of the process may be made open through scholarly reporting
and writing on teaching practice, some aspects may
be informally discussed with fellow teachers, and another view of what transpired may be made open via
student feedback and ratings. However these are all
interpretative events — it is not possible for interested
parties to gather primary data on what happened in a
particular teaching situation unless they happened to
be physically present at the time. In our experience
teaching practice is generally unobserved.
Yet openness and genuine collegiality between
teachers, open discussion and critical reflection on
teaching practices has well established and significant impact on the development of teachers and on
the quality of their teaching (see for example Patricka
et al. (2010) and Daya (1993)). Indeed open practice
and a concomitant culture of review and reflection is
the hallmark of a heathy profession.
When I started my career I really didn’t
know how to teach maths. I didn’t have the
training for it. So, in the beginning I had to
find out how to teach it. I needed some help.
If you’re only a beginner, it’s difficult to get
some help. Also, my colleagues were very
busy. They didn’t have the time to help me.
So I had to solve my problems on my own.
(Clement & Vandenberghe 2000)
There are a number of established ways of teachers addressing this issue and sharing their practice.
Two of the most widespread are peer review and team
teaching. These are both rich methods with being investigated by an active research community and provide the potential for a range of benefits when they
are used sensitively in an environment of trust under a
developmental rather than judgemental ethos (see for
example (Gosling 2005), (Gosling & O’Connor 2009)
and (Anderson et al 2009)). Here we simply note that
these approaches provide environments where it is acceptable for teachers to observe the teaching of their
peers. Openly publishing lectures online provides another such environment, with the potential to share
some of the same benefits, and likely subject to similar mitigating factors.
My own teaching practice has been improved significantly over the years by observing other teachers
teaching. For the first five or six years of my career
as a university lecturer I made a habit of attending
and observing the lectures of other teachers at local
universities about whom I had heard students speak
highly. This had a noticeable effect on the effectiveness of my own teaching (indeed I subsequently received a number of teaching awards). However the
observation process was time consuming and awkward
to arrange. The current practice of university teaching is a largely private and closed arrangement. Our
students know how we teach, but in the large our
peers and colleagues do not.
On the other hand it is easy to “visit” the classrooms of those who publish their lectures online. By
publishing we open the doors of the classroom and let
in the light.
3.2
Open Learning
Our second “lofty goal” was open learning — to disseminate our teaching so that it is available to anyone who might wish to learn from it. For example this
might include students at our own university, students
at other universities, students not yet in university,
those unable to attend university for socio economic
reasons, or reasons of health or time constraints, and
those already in a career looking for ongoing professional development or contemplating a career change.
For me this was what education is about - humanity striving to improve itself, thirsting for knowledge
and to know more. Putting university teaching online is reminiscent of the transformative power of the
Children’s Television Workshop back in the 1960s and
1970s, where open freely given knowledge and teaching gave new opportunities to inner city children with
limited educational options (Lesser 1974).
4
Approach
This section sets out in detail the process we use to
record and publish the lectures. The specific details
of the makes and models of the equipment we used is
set out in the Appendix. Our hope is that this will
be a useful resource for others interested in dipping
their toes into the water.
In general my experience has been that most of the
things that I worried about beforehand turned out to
be remarkably easy to solve in a satisfactory manner.
These included:
1. What hardware to use?
2. What level of quality to aim at?
3. How to publish on the internet?
4. Intellectual property?
5. Will it be too expensive?
6. Will it even work, at all?
7. Fear of looking like an idiot in front of the whole
world, irrevocably...
On the other hand the main challenges which
arose, and which have still not been solved perfectly,
were things I had not anticipated:
1. Correspondence can consume vast amounts of
time (and likely contributed to my developing
RSI in late 2008)
2. How can we have the recording work every time,
without fail?
3. How can we set up the gear and pack away the
gear in time?
4. Hateful comments on YouTube can tear your
heart
4.1
Selection of Lecture Theatre
A critical requirement for the recording process was
that it not diminish the experience for the students
sitting in the room. At the end of the day they were
most important aspect of the lecture and I didn’t want
them to feel like a studio audience, I didn’t want the
students in the lecture theatre to feel that the recording was the main thing and they were secondary. We
decided to locate all the cameras and recording gear
at the back of the room, so the camera and camera
operator did not visually intrude during the lecture.
There were only a few lecture theatres which were
both large enough for our class (about 250 students)
and in which the back row was not too distant from
the front of the lecture theatre to record clearly in ambient light. Essentially we needed large, shallow but
wide theatres. Furthermore the camera crew had to
get themselves and their gear in and out in the brief
(10 minute) gap between successive lectures. Luckily we were able to find and get permission to use
a wide theatre with a separate back entrance which
we could mark off with portable signage and reserve
for the camera crew without obstructing safe exit in
case of fire or other emergency. The distance from
the camera position to the blackboard was about 15
meters.
4.2
Keep it simple
We did a number of dress rehearsals before the start
of semester and soon discovered that the biggest challenge was set-up and pack-up complexity and speed.
Our hard limit was the length of the interval between
the previous lecture finishing and the lecture to be
recorded starting. This meant we had to be reliably
set up and be ready to roll in ten minutes. This became the limiting factor in our recording design. Basically there was no point in having some fantastic
but complex piece of equipment if it took so long to
set it up that we missed capturing the start of the
lecture.
4.3
Cameras and Microphones
I wanted to record the output from slides/data projector/blackboard/document camera and also to record
the lecturer and to see the backs of the heads of
students to give viewers the effect of being in the
room rather than the effect of watching a slick AudioVisual presentation. The decision to adopt a Cinéma
Vérité approach had the fortunate consequence of allowing us to use consumer quality recording equipment which was considerably cheaper than professional equipment.
For recording the lecturer we used a high end consumer grade HD digital video camera. It was a common camera so we were easily able to find a spare
to swap when the camera went in for its one repair
(tip: never plug firewire cables in backwards — you
can do it without too much force and it reverses the
power supply and blows the camera and the firewire
board on the computer. We now use an in-line polarity protecting electronic fuse...). We mounted the
camera on a professional tripod with fluid filled head
to allow steady recording and smooth panning. For
audio the lecturer wears a wireless microphone and
transmitting pack, and the receiver is mounted on
the camera. The camera we used accepted multiple
audio inputs which was very convenient as it permitted us to mix in the microphone audio channel on the
camera itself and not have to worry about an extra
data stream and extra cables to the computer.
We tried various clever hardware and software approaches to record the images from the backboard and
projected from the data projector/document camera
etc. Luckily we were able to try out proposed solutions before having to pay much money as these did
not live up to our expectations. The two general types
of approaches we initially tried were:
1. Resampling the SVGA signal going into the data
projector (quality too low, didn’t capture the
backboard, needed slow moving facilities staff to
implement a hardware solution for each room we
used before we could record in it);
2. Vodcasting software to capture the screen from
lecturer laptop (quality too low, didn’t capture
black board, document camera, or the built-in
computer in the lecture theatre, needed a long
cable run from front of the theatre to the mixing
area at the back)
None of these were reliable and flexible enough for
our requirements. So we instead adopted the naive solution of purchasing a second camera (identical to the
first) and simply pointing it at the blackboard/screen.
That meant we could capture data projected from any
source. Furthermore for reliability it meant we were
physically isolated from, and so not dependant on,
the existing (unreliable) AV equipment in the lecture
theatre. For speed of setup we didn’t even attach
this camera to another tripod — since it was fixed
in position for the duration of the lecture we simply
placed it on a sandbag and pointed it in the right
direction. At 15 meters with near maximum optical
zoom it produced good quality video.
We attached a high grade microphone to the fixed
camera as a backup in case the lecturer’s radio mike
failed, and to capture ambient sound in the room such
as questions from students.
4.4
Audio Reliability
The camera operator always wore headphones
plugged into the camera they were operating so they
would know instantly if there was a problem with the
sound. We used fresh batteries in the radio mike and
receiver each time. This seemed wasteful but after
once experiencing losing sound due to flat batteries
we never wished to risk that again!
Periodically I offered the resultant half-used AA
batteries to students who seemed excited to be offered
a gift of such riches. Perhaps they’ll recall this and endow the university with similar generosity when they
are running Google...
4.5
Mixing and Data Logging
Both the technicians in the University’s own Audio
Visual unit, and the two professional video editors I
initially consulted for advice used the same editing
software and the same family of hardware so we simply mimicked their setup. No doubt a vast range of
other configurations would also have been satisfactory for our purposes but we didn’t have the time
or tolerance for possible recording problems to permit the luxury of experimentation. Working on the
same system as the locally available and friendly experts seemed the most prudent approach to adopt to
minimise risk of problems arising which could not be
solved rapidly.
We choose to capture on a desktop machine rather
than a laptop. It would have been more convenient to
capture and mix on a laptop but the laptops we experimented with all had problems processing 3 firewire
streams in parallel at high data rates and we wished
to have the capability to use 3 cameras if needed (to
date we have never needed 3 cameras however). In
addition the desktop machine has 2 quad core processors which provided much faster postproduction
editing and processing than a laptop.
Mixing and processing audio visual data is astonishingly time consuming. A very minimal edit takes
me about five minutes per one minute of footage. This
means five hours for a one hour lecture. This was not
sustainable in our context as I was already very busy
lecturing and running the course and could not find
an extra 10-20 hours time to work on video processing each week. Furthermore we needed to release the
videos to the off campus students within a day or so.
And this fast turnaround would likely be most critical precisely when the course, and consequently I,
was most busy (near assignment due dates etc.) After much consultation and brainstorming we decided
the most reliable strategy was to mix the footage live
during the lecture (much like a television studio does)
and post the already mixed footage immediately at
the end of each lecture.
We used live mixing software which could be preset with common mixes of the two cameras (e.g. camera A only, camera B only, camera B with the output
from camera A picture-in-picture, and so on) and fade
between the presets at a mouse click. In theory the
camera operator probably could have also done the
mixing at the same time but for reliable setup, packing up, and dealing with contingencies we had usually
had two operators at the back, one mixing and one
operating the camera.
The two cameras were connected to the capture
computer via firewire, they were mixed live and the
resultant stream was saved to the hard disk in real
time. We also kept DV tapes running in each camera
in case of catastrophic or in case we wanted to re mix
the streams afterwards. There have been a few times
we have needed to remix from these backup tapes
after the lecture as occasionally the mixing software
freezes or drops a camera, and for example one time
the power cord was kicked out from the computer.
The remixing in these cases has taken a considerable
amount of time and although the editing is marginally
slicker when planned rather than performed live the
benefit is far outweighed by the extra time cost.
4.6
Hosting and Uploading
There are a range of ways to host and distribute Open
Educational Media (OER) such as lecture recordings.
They are not mutually exclusive - we used several.
The most commonly used approach currently
seems to be to host the media files on servers at the
host university, usually in an ad hoc fashion schoolby-school or even course-by-course. Advantages of
this approach is it is easy to arrange and monitor
and other course material can be integrated.. The
main disadvantage is that, except for local students
with whom you can directly communicate, it can be
hard for others to find the material. This is largely a
push model of distribution. The most successful exemplar of this approach is MIT who provide lectures
and course material via their dedicated OpenCourseWare portal (MIT-OCW).
To allow others to find the material more easily
it can be submitted to well known learning repositories such as Academic Earth, the OpenCourseware
Consortium, YouTubeEDU, and iTuneU.
We hosted the video files on our school’s local fileserver for local students but in order to make the
material easily accessible globally we also published
the videos on YouTube and on iTunesU. Uploading to
YouTube is easy but slow. Some of our recent uploads
have taken over 5 hours per lecture.
Two benefits of YouTubeEDU are that Google
hosts the files for you at no cost (and with no maintenance effort on your part), and that there is already
an active viewer community on YouTube so the videos
are easily accessible to an existing and substantial audience. Uploads to YouTube have recently been permitted to exceed the previous 2GB limit per video,
and educational users are permitted to upload videos
of any duration (normal YouTube uploaders are limited to videos of no more than 10 minutes.) In practice about 2GB is sufficient to provide a high quality
HD video of a one hour lecture at 720p resolution.
YouTube provides a range of analytical tools to allow
posters to see where their viewers are coming from,
and also provide audience demographic (age, gender)
data when it is known. They also provide clear and
detailed data on how your viewers have discovered
your videos — which is invaluable in selecting and
adjusting keywords and video titles.
An advantage of iTunes is that you host the files
yourself and so have full control, a possible disadvantage is that you have to organise and pay for hosting
and bandwidth. Academic Earth has the advantage
of being able to host course material as well as lecture
videos.
Be aware that the YouTube uploading process is
quite fragile, accidentally navigating away from the
upload page during upload aborts the whole process,
as can engaging in other YouTube related activity it
seems. YouTube offer a bulk uploader but I have
sometimes found this to be unreliable and have got
better results uploading the videos manually one at a
time. I upload from a dedicated machine and don’t
touch the machine again until the upload has complete since accidentally aborting a 5 hour upload after
4 hours is very frustrating. Uploads also fail erratically on completion with terse error messages suggesting that the video file is in the wrong format, trying
again with an identical file often works.
Viewers in YouTube can interact with the video,
they can add annotations, write comments, rate the
video, add it to their own public compilations, and
recommend it to their friends. As a teacher I find
the feedback data provided by YouTube is useful in
improving the effectiveness of the videos, in contrast
to the basic hit count data our University provides to
those hosting vodcast files on the University’s iTunes
server. I also like the (admittedly limited) sense of
community amongst the YouTube viewers as opposed
to the solitary experience offered to my iTunesU viewers.
4.7
IP
Our lectures and course material are released under
the Creative Commons Attribution-NonCommercialShareAlike 3.0 licence. This means copyright of the
lecture itself remains with the lecturer but anyone is
permitted to copy, remix and derive further works
from the material for non commercial purposes provided they provide attribution and impose the same
conditions on the copies and derived works.
5
Outcomes
This section first reports on the outcomes of the
project with respect to our two initial objectives for
recording the lectures. It then reports on the outcomes of openly publishing, whose significance I feel
far eclipses the original aims of the project.
5.1
Secondary School Outreach Programme
The secondary school outreach program has eventually proved a substantial success although the 2008
pilot involving three students had a number of problems. That year one student dropped out, and the
other two had problems accessing the videos online
due to the bandwidth limits of their home internet
connection. Furthermore we had problems uploading
the videos to YouTube in a timely manner - largely
as a result of the problems outlined in the previous
section.
In 2009 the program ran again with the already
recorded 2008 lectures, this time with 16 students. By
this time all of the lectures were already on YouTube
but a few students also reported problems with their
download quotas. All the files were available for free
download over the university wireless network which
mitigated this problem to some extent. In 2010 we
provided all students with the video files on DVDs so
no downloading was required.
The response of the students taking the course
has been overwhelming. Although we provide formal
transcript qualification at the end of the course suitable for credit at any university three of the four students who have completed secondary school to date
have come to study computing at the university —
suggesting that they found the experience rewarding.
Many of the other students from the course, currently
in their final year of secondary schooling, have stated
a firm desire to proceed to university and continue
to study computing. The level of enthusiasm, excitement, and community amongst the graduates of
the course is quite remarkable. At the 2009 UNSW
open day a number of the secondary students turned
up at the university of their own initiative and organised their own stall where they demonstrated the
project they were currently working on. We observed
two of the students speaking excitedly to visiting students about the fun of studying computing at university. Students from the 2009 course have subsequently
volunteered to tutor and mentor the 2010 students.
This shows considerable dedication from them as at
the same time they are studying for their own HSC.
Interestingly approximately half of the returning tutor/mentors are female, which contrasts with our typical first year intake of around 20% female students.
If this trend persists it will warrant investigation.
On average the academic performance of the 2009
students who watched video lectures and face to face
tutorials and labs exceeded the performance of the
students who physically attended lectures and had
face to face tutorials. Obviously the sample size is
small and the group was positively selected so the two
groups are not directly comparable but it does not appear that the students who did not have face to face
lectures were at any significant educational disadvantage. Indeed the video students themselves reported
high levels of satisfaction with the way the course
was organised, the average response to the summary
question “Overall, I was satisfied with the quality of
this course” was 4.8/5.0 (15 responses) for those who
watched lecture videos, compared with 4.4/5.0 (184
responses) for the university students who did physically attend lectures. These results are comparable to
those experienced in CMU’s Open Learning Initiative
(Lovett et al 2008).
The high positive response from the videowatching students may well have been a consequence
of the resultant shift in focus from lectures to the
group based tutorial-lab sessions. These students
watched lectures at home privately as preparation for
the tut-lab session, which was then seen as the focus of the course each week. Being a tutorial-lab it
was highly interactive and personalised perhaps making the overall experience seem much more social and
personally relevant. We are currently gathering data
from the 2010 course participants to further investigate this hypothesis.
5.2
Sharing teaching practice, and Syllabus
Stability
We are still working on the project of using the recordings to share ways of teaching higher level content
such as group work, skepticism etc. Some teachers outside of the group we were targeting with the
lecture recordings have unexpectedly approached us
and said they have watched some and in some cases
all of the lectures and have found them useful in
their own teaching practice. From informal conversations it seems many of the teachers in the school
have watched at least one of the lectures but it is
extremely unlikely that many have had the time or
inclination to watch the full 50 hours for any course.
We are currently working on a teacher summary for
each course which gives links to small fragments of
video demonstrating each of the more novel aspects
of the current syllabus. This is a work in progress and
is publicly available for any interested teachers (see
https://wiki.cse.unsw.edu.au/openlearning).
One interesting aspect of having lectures captured
as open and persistent objects is that this makes them
available as first class objects for academic discourse
and analysis. For example in a paper I am currently
writing on the effective teaching of skepticism I am
able to cite specific moments in various lectures to illustrate points — and then readers are able to see the
teaching at these moments for themselves rather than
just reading my second hand reports of my own interpretation of what I recall doing. Further they can
roll forward or backwards to any other part of the
lecture, or indeed view preceding or subsequent lectures, and investigate to their own satisfaction issues
of context and acts of summary and selection I perhaps unknowingly perform when reporting the work.
This is consistent with the emerging open data movement discussed in Section 3.0 above, and parallels the
modern practice in scientific scholarship of making
raw data publicly available in addition to the conclusions drawn from the data. This approach offers many
novel and exciting possibilities and suggests that dramatic changes may lie ahead in conducting and disseminating the scholarship of teaching and learning.
5.3
Unanticipated Impact
As reported above the lecture videos have been
watched over one million times. They have been
highly rated by viewers on YouTube all being rated
either 5 or 4.5 stars. One of the lecture videos was
selected by YouTube as a Featured Video on the
YouTube front page world wide. The response from
students and teachers worldwide has been somewhat
overwhelming and by late 2008 responding to emails
and written letters about the lectures was taking
hours each week. Many of these interactions were
from university students currently studying computing at other institutions but signifiant impact of the
lectures was on those for whom university was not an
option. I received messages from viewers were from
third world countries without local university options,
from viewers who were not permitted to attend university, from those who could not afford university,
from younger students bored at school but not yet of
university age.
THANKYOU for posting [your lectures]. I
am dyslexic with the reading comprehension
of a 7yo and the language ability of a 35yo,
have pretty much fallen behind at uni within
the first 3 weeks of every course (at [another
university] not UNSW). Putting the lectures
online is a real help as it gives people the
chance to learn at their own pace instead of
under pressure of the course. It is definitely
a worthwhile project, and a great supplement.
YouTube user Battlewench comment on UNSWelearning YouTube channel (Battlewench 2009)
Existing mechanisms for course feedback and assessing the effectiveness of my face to face teaching
do not translate in any obvious way to assessing the
learning arising from the largely anonymous group of
remote learners watching the material online. I am
currently engaged in research project to investigate
possible ways to address this challenge. Nonetheless
the flood of qualitative feedback I have received since
starting to publish the lecture recordings makes it
clear that the project has had an educational impact
dwarfing all my previous work. This has been both
humbling and inspirational and is incredibly motivational as a teacher.
The online lectures currently average 1,500 viewers per day watching 3,000 videos. It is hard to divide
these viewer numbers into students and fellow teachers. An interesting indication is that Google reports
70% of our viewers are over the age of 34. I have
received considerable correspondence and YouTube
comments from other teachers who have enjoyed
watching a different perspective on teaching. As an
indication the most recent teaching comment from a
viewer today is quoted below.
I look forward in turn to observing the teaching
practice of other teachers who engage in open teaching. Current practitioners of open teaching in computer science include Mehran Sahami, Eric Grimson
and John Guttag. I believe that teachers openly sharing our practice in this way will improve the teaching
profession and the quality of what we do.
what Clement & Vandenberghe (2000) refer to as
strategic autonomy in order to avoid the criticism
of colleagues. This may well prove the most significant barrier to widespread adoption of open
teaching.
Like for instance, if you have difficulties with one of the children, the
real and problem children, then it’s
delicate to discuss that with your colleague. You don’t feel inclined to
tell about it, because the other would
doubt your competence. Especially if
you’re a novice teacher. Then they
look at you with Argus eyes, don’t they
— teacher comment (Clement & Vandenberghe 2000)
Wow!! Our high school here in Texas has
gone to a 1:1 laptop to student setup. We
are in our 3rd year of computers this year.
I thought I had exhausted the possibilities!
Now I see that I’ve only begun... I can’t wait
to try the wiki experience this year with my
students. Thank you..!! — (SkyRookie1 2010)
6
Issues
After the mechanical details of recording and publishing lectures are resolved there remain important
questions about how to make the process effective for
learning. We set out briefly issues which need to be
carefully considered.
• Consent: Students present in the lectures must
not only consent to the recording but the act of
recording and perhaps their concerns about being recorded must not interfere with their own
learning. Our approach is that the students in
the class are our primary concern, the recording
is a nice bonus if it works. We consult with students in the first lecture, we strive not to record
faces of students, we identify areas they can sit
to avoid being recorded, and we ask that if anyone is unhappy with their comments in a lecture
being recorded that they advise us as soon as possible after the lecture and we will erase this before posting the footage. Despite providing these
avenues for students to exclude themselves from
the recording to date no student has raised any
concerns. This needs further investigation but
perhaps they are simply not as concerned with
privacy as say my generation is.
• Engagement: watching videos is passive by nature, and one hour lectures are much longer than
the 10 minute norm most viewers have become
accustomed to on YouTube. To stop viewers
falling asleep we use a combination of active camera work, cuts between cameras, and the way the
lecture is structured in an effort to keep the students engaged. As mentioned above we strive
to make experience for viewers feel like they are
present in the class rather than simply watching
a passive recording of it. Luckily also the discipline of computing is very interesting in and of
itself!
• Criticism: Comments on the internet can be
made quickly and with little thought. Debate can
be robust. Furthermore such comments are made
publicly and are available for all to see. I follow a policy of only deleting comments which include offensive language or which are disrespectful to specific students in the audience, and leave
other critical comments untouched. Some of the
comments and feedback I have received on my
teaching has been quite confronting. It seems
likely that some teachers may not wish to be exposed to blunt and public criticism of their teaching. Furthermore it is possible that some teachers will not wish to share their teaching even if
public commenting is not permitted, engaging in
7
Conclusion
Publishing videos of my lectures online has proved
to involve only a modest amount of work and virtually no cost, despite the university having no useful recording infrastructure in place. This was made
possible due to supportive colleagues and university
management, enthusiastic student volunteers, and the
adoption of a Cinéma Vérité approach and live mixing.
From my point of view the main costs have been
the time taken to respond to correspondence from
remote learners and the continual guilt I feel for not
being able to do this adequately.
The rewards have been unexpected and immense,
both emotionally and in the impact I have been able
to have with my teaching on students worldwide and
on other teachers. I are excited by the potential for
video recording to transform the practice of teaching
from something that happens largely privately and
behind closed doors into an open and communal process.
8
Future Work
Further work is needed to develop useful ways of assessing and improving the effectiveness of teaching
and course design when learners are largely anonymous; have widely differing and unknown levels of
prior knowledge, skills, and learning objectives; and
need not follow the course in a linear fashion.
9
Acknowledgements
I acknowledge and thank my students from
COMP1917 (2008 semester 1), COMP1927 (2009
semester 2), and COMP2911 (2010 semester 1) for being so supportive of our time together being recorded,
and for not letting the presence of the cameras dim
their enthusiastic and active engagement during lectures. Our courses are based around a strong sense
of community. If the classes look fun and interesting, it’s because they were, and this arose from the
community we formed while each course unfolded with students contributing at least as much to the
mood of the class as I did as lecturer. It was our
class, not my class, and it was with great generosity
and open heartedness that the students were willing
to allow outsiders to share in their experiences. Furthermore I thank my students for their unquestioned
trust, which I found quite moving, that I would treat
them respectfully in the editing and release of the
recordings of our times together.
That this paper was written is due in large part
from the encouragement and support I have received
from my many wonderful colleagues with a passion
for education. Most directly in the case of this paper
Helen Dalton, Jan McLean and Judy Kay who asked
great questions, and Matthew Clarke who provided
helpful writing advice. Matthew also very kindly read
and commented on an earlier version of the paper
and this version has been improved by a number of
his comments. Likewise the paper has benefited substantially from the critiques and insights contributed
by each of the three anonymous reviewers. I thank
you all most warmly for the care you put into making
very helpful suggestions.
I would like to acknowledge all those who by their
encouragement and efforts made the lecture recording possible. At the time we started there were few
widely known exemplars of public videos of university
courses outside of MIT-OCW (iTunesU and India’s
nptelhrd channel had only just started, in the case of
iTunesU most content was audio only) and certainly
none locally, and it would have been very easy for
someone somewhere to have said “no, it’s too hard”
or “it won’t work.” Eliathamby Ambikairajah, the
pioneer of lecture recording at UNSW, gave me much
useful information about the impact of recording on
students in the classroom and watching his recordings
provided reassuring evidence that current camera resolutions were sufficient and that it could be done.
Patrick Stoddart first suggested the idea of making the recordings public, Mary O’Malley, Tom Cavdarovski, Mark Foster, Michael Rampe and Daniel
Woo provided wise and practical advice on the mechanics of video recording, Paul Compton and Ashesh
Mahidadia first thought of the idea of live mixing.
David Collien and Theo Julienne were constant advocates for resuming outreach to show bored secondary
school students the joy of computing in the same way
they themselves had been inspired by the CSE computing club in the 1990s.
The biggest contributions of any individuals to the
project was the heroic and entirely voluntary (they
repeatedly refused payment) efforts by Rupert Shuttleworth and Thurston Dang, former students of the
courses, who operated the cameras and live-mixed the
lectures throughout the project. They set up and
packed up the gear, sorted out hardware and software problems, and just generally made everything
happen. Their efforts meant I never had to worry
about recording when I needed to be worrying about
the lecture. Furthermore they are intelligent and effective educators. I deeply valued our times travelling
back together to my office after lectures to put away
the recording equipment when we could talk about
the course together, conducting a postmortem on the
lecture that had just happened, and bouncing ideas
around for what was still to come. Being open in my
teaching practice and the consequent benefits from
Thurston and Rupert being able to observe what had
happened in lectures and discuss it in a rich manner
was a compelling endorsement of open teaching.
Appendix: Equipment List
Camera: 2 x Sony A1 HDV
Tripod: Miller + DS-5 head
Wireless Microphone: Sony UTX B1 / URX
B1
Computer: Apple Mac Pro 2x2.4GHz QuadCore
Live Mixing: WireCastHD by Telestream, Inc.
Postproduction: Adobe FinalCut Pro
References
Anderson, L., Krathwohl, D., Airasian, P., Cruikshank, K., Mayer, R., Pintrich, P., Raths, J. & Wittrock, M. (2001)A Taxonomy for Learning, Teaching, and Assessing A Revision of Bloom’s Taxonomy of Educational Objectives, Addison Wesley
Longman.
Anderson T. D., Parker N. & McKenzie J. (2009)Assessing online collaboratories: a peer review of
teaching & learning ATN Assessment Conference
2009, RMIT University.
Battlewench (2010), YouTube user, Comment
on UNSWelearning channel, comment posted
early 2009 (youtube does not show precise
date). Accessed at comment page 9 (of 14) on
http://www.youtube.com/user/UNSWelearning
on 4 November 2010.
Bloom B. S. (1956). Taxonomy of Educational Objectives, Handbook I: The Cognitive Domain. David
McKay Co Inc., New York.
Britzman, D. P. (1986), Cultural myths in the making of a teacher: Biography and social structure in
teacher education, in Harvard Educational Review,
56(4), pp. 442–456.
caesiume (2010), YouTube user, Commenting on
video Lecture 1: Introduction to Data Structures and Algorithms - Richard Buckland,
comment posted January 2010. Accessed at
http://www.YouTube.com/comment_servlet?all
_comments=1&v=RpRRUQFbePU on 5 August 2010.
Clement, M. & Vandenberghe, R. (2000), Teachers’
professional development: a solitary or collegial
(ad)venture?, in Teaching and Teacher Education,
16, pp. 81–101.
Daya, C. (1993), Reflection: a necessary but not sufficient condition for professional development, in
British Educational Research Journal, Volume 19,
Issue 1, pp. 83–93.
Gosling D. (2005), Peer Observation of Teaching.
SEDA Paper 118: Staff & Educational Development Association Ltd, London.
Gosling D. & O’Connor K. eds (2009), Beyond the
Peer Observation of Teaching. SEDA Paper 124:
Staff & Educational Development Association Ltd,
London.
Lesser, G. (1974). Children and Television: Lessons
from Sesame Street. Random House, New York.
Lovett M., Meyer O. & Thille C. (2008). JIME - The
Open Learning Initiative: Measuring the Effectiveness of the OLI Statistics Course in Accelerating
Student Learning in Journal of Interactive Media
in Education, May 2008, 14.
OECD (2004). Science, Technology and Innovation
for the 21st Century. Meeting of the OECD Committee for Scientific and Technological Policy at
Ministerial Level, 29-30 January 2004 - Final Communique.
Patricka, F., Elliota, D., Hulmea, M. & McPheea,
A. (2010), The importance of collegiality and reciprocal learning in the professional development
of beginning teachers, in Journal of Education for
Teaching, Volume 36, Issue 3, pp. 277–289.
Rogers D. L. & Babinski L. M. (2002). from isolation
to conversation: supporting new teachers’ development. SUNY Press, New York.
SkyRookie1 (2010), YouTube user, Commenting
on video Wikis in University Teaching and
Learning - Richard Buckland UNSW, comment posted 15 August 2010. Accessed at
http://www.youtube.com/watch?v=m1-8OOrBi0o
on 15 August 2010.
Uhlir, P. F. & Schroder, P (2008), Open Data
for Global Science, Sydney University Press Law
Books 39 in Brian Fitzgerald, Legal Framework for
e-Research: Realising the Potential 188.
Vest, C. M. (2001), quoted in MIT to make
nearly all course materials available free on
the World Wide Web MIT news April 4 2001
http://web.mit.edu/newsoffice/2001/ocw.html
(accessed 4 November 2010).